Intelligent station emergency joint defense system
By using multi-camera collaborative positioning and AI video analysis modules, combined with UWB positioning and blockchain encryption technology, the problem of isolated operation of fire protection and medical systems in smart stations has been solved, enabling efficient and accurate emergency response within smart stations.
Patent Information
- Application Number
- CN202510934941.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-08
- Publication Date
- 2025-10-28
AI Technical Summary
In existing smart stations, subsystems such as fire protection and medical services operate in isolation, rely on manual dispatching for coordination, and suffer from slow sensor response and coordination deficiencies.
It employs a multi-camera collaborative positioning module, an AI video analysis module, an emergency response module, and a data security module, combined with deep learning algorithms, UWB positioning technology, and blockchain encryption, to achieve seamless cross-regional tracking, automatic emergency response, and data security.
It enables continuous target tracking and rapid emergency response in complex environments, improving the efficiency and accuracy of emergency response coordination and reducing the impact of information gaps on rescue efforts.
Smart Images
Figure CN120852121A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the technical field of emergency management systems, and more specifically, to a smart station emergency joint defense system. Background Technology
[0002] Traditional station monitoring relies on manual shifts, making it prone to oversights due to visual fatigue. Smart stations, building upon existing digital and intelligent stations, fully leverage next-generation technologies such as artificial intelligence, big data, cloud computing, AIoT, and digital twins. They provide passengers with a comprehensive experience, offer intelligent operation and maintenance data support, provide panoramic control for station operations, and offer decision support for management. This achieves safer operations, smarter services, and more efficient management by developing a smart system encompassing five key areas: holographic perception, intelligent analysis, panoramic control, precision and convenience, and proactive evolution.
[0003] However, existing smart stations also have problems such as isolated operation of subsystems such as fire protection and medical care, reliance on manual dispatch for linkage, slow sensor response, and various coordination defects. Based on the above problems, we have provided a smart station emergency joint defense system. Summary of the Invention
[0004] To address the problems mentioned in the background section, this invention provides an intelligent station emergency joint defense system.
[0005] The intelligent station emergency joint defense system provided by this invention adopts the following technical solution: The intelligent station emergency response system includes: The multi-camera collaborative positioning module consists of distributed high-definition positioning cameras, which use visual algorithms to achieve seamless cross-regional tracking and three-dimensional coordinate positioning of targets. The AI video analysis module uses deep learning algorithms to identify abnormal behaviors such as falls, gatherings, and leaving items behind by people in the station in real time, and outputs the type of abnormal behavior and the coordinates of its occurrence. The emergency response module includes a digital emergency plan library, which automatically triggers associated fire alarms, power control, and evacuation guidance linkage operations when responding to abnormal events. The data security module uses AES-256 (Advanced Encryption Standard - with 256-bit key) encryption technology to encrypt and store personnel information, and sets up a 24-hour automatic deletion mechanism based on timestamps.
[0006] Preferably, the multi-camera collaborative positioning module includes an infrared thermal imaging unit, which continuously tracks the target through thermal radiation characteristics in low-light environments, with a thermal imaging resolution of 384×288 pixels. It can also add a multispectral fusion algorithm to combine data from multiple sensors such as visible light, infrared, and lidar to improve the robustness of target recognition.
[0007] Preferably, the multi-camera collaborative positioning module further includes a trajectory prediction unit, which predicts potentially dangerous behaviors by analyzing personnel movement paths.
[0008] Preferably, the AI video analysis module integrates an occlusion compensation algorithm, which enables continuous tracking when the target is occluded through spatiotemporal context association. The specific steps are as follows: 1) Video surveillance detects that the target is occluded, and predicts the target's trajectory through spatiotemporal context modeling; 2) The UWB positioning module continuously provides the coordinates of anchor points in the surrounding environment, and uses the Kalman filter algorithm to fuse visual prediction results with real-time UWB data; 3) Generate a compensated continuous coordinate flow.
[0009] Preferably, the emergency response module includes a medical emergency submodule and a fire response submodule. The medical emergency submodule automatically obtains the coordinates of the nearest AED device through UWB positioning technology and directly connects to the hospital's emergency department. The specific steps are as follows: 1) When abnormal behaviors such as falls are detected, a health data request is triggered, and the passenger's identity identifier is quickly read through RFID / NFC tags; 2) Blockchain nodes verify identity and permissions, based on the Hyperledger-Fabric channel isolation mechanism, and smart contracts are automatically approved: rescue personnel need to pass two-factor authentication; 3) Decrypt and push health information; data transmission uses TLS 1.3 + SM4 encryption (Chinese national standard). 4) After the emergency response is completed, the event record is written to the blockchain, which includes timestamp, location coordinates, data summary, and response result.
[0010] Preferably, the fire-fighting submodule includes a lidar fire source location unit, which can generate three-dimensional fire data and link with the fire pipeline control module. After a fire occurs, it automatically shuts off the power in the fire area and generates three-dimensional fire location data.
[0011] Preferably, it also includes a special passenger service module, which contains a biometric comparison unit that matches the identity information of special passengers through facial recognition.
[0012] In summary, the present invention has the following beneficial technical effects: 1. This application adds a cross-domain technology integration of "occlusion compensation algorithm + UWB positioning + blockchain health data" to maintain good tracking continuity in complex scenarios such as occlusion and low light. By using a high-definition positioning camera equipped with AI vision algorithm, it can track the movement trajectory of people in the station in real time; it significantly improves the target tracking continuity in complex environments. By using an occlusion compensation algorithm combined with UWB positioning technology, it can achieve continuous tracking through motion trajectory prediction when the target is completely occluded by luggage or crowds.
[0013] 2. This application introduces a multimodal perception system of "thermal imaging tracking + three-dimensional fire modeling". When a fire is caused by an electrical short circuit, the system automatically triggers the "fire plan", simultaneously activating the fire alarm throughout the station, shutting off the power in the fire area, sending the precise fire location to the fire department, and guiding the intelligent signs in the station to switch evacuation routes to quickly evacuate the public.
[0014] 3. In this application, by actively capturing abnormal passenger behavior, even if a passenger loses consciousness and is unable to call for help, the system can detect the abnormality immediately, display the specific location in real time, and immediately initiate rescue, achieving "responding before being called". While issuing an early warning, the system will also push the passenger's medical history information in real time, so that rescuers can have a more comprehensive understanding of the situation and take targeted emergency measures, avoiding the impact of missing information on the treatment effect. Compared with traditional indiscriminate rescue, it is more accurate and efficient. Attached Figure Description
[0015] Figure 1 This is a block diagram of the overall system architecture of the present invention; Figure 2 This is a system diagram of the occlusion compensation algorithm of the present invention; Detailed Implementation
[0016] The following is combined with Figures 1 to 2 The present invention will be described in further detail below.
[0017] It should be noted that the accompanying drawings are schematic and not to scale. For clarity and convenience, the relative dimensions and proportions of the parts shown are exaggerated or reduced in size; all dimensions are merely illustrative and not limiting. Furthermore, the same reference numerals are used for the same structures, elements, or fittings appearing in more than two drawings to indicate similar features.
[0018] This invention discloses a smart station emergency joint defense system, comprising: The multi-camera collaborative positioning module consists of distributed high-definition positioning cameras. It uses visual algorithms to achieve seamless cross-regional tracking and three-dimensional coordinate positioning of targets, with a positioning accuracy error of ≤0.3 meters. The AI video analysis module uses deep learning algorithms to identify abnormal behaviors such as falls, gatherings, and leaving items behind by people in the station in real time, and outputs the type of abnormal behavior and the coordinates of its occurrence. The emergency response module includes a digital emergency plan library. When responding to abnormal events, it automatically triggers associated fire alarms, power control, and evacuation guidance linkage operations with a response delay of ≤2 seconds. The data security module uses AES-256 (Advanced Encryption Standard with 256-bit key) encryption technology to encrypt and store personnel information, and sets up a 24-hour automatic deletion mechanism based on timestamps.
[0019] Preferably, the multi-camera collaborative positioning module includes an infrared thermal imaging unit, which continuously tracks the target through thermal radiation characteristics in low-light environments, with a thermal imaging resolution of 384×288 pixels. It can also add a multispectral fusion algorithm to combine data from multiple sensors such as visible light, infrared, and lidar to improve the robustness of target recognition.
[0020] Preferably, the multi-camera collaborative positioning module further includes a trajectory prediction unit, which predicts potentially dangerous behaviors by analyzing personnel movement paths.
[0021] Preferably, the AI video analysis module integrates an occlusion compensation algorithm, which enables continuous tracking when the target is occluded through spatiotemporal context association. The specific steps are as follows: 1) Video surveillance detects that the target is occluded, and predicts the target's trajectory through spatiotemporal context modeling; 2) The UWB positioning module continuously provides the coordinates of anchor points in the surrounding environment, and uses the Kalman filter algorithm to fuse visual prediction results with real-time UWB data; 3) Generate a compensated continuous coordinate flow.
[0022] Preferably, the emergency response module includes a medical emergency submodule and a fire response submodule. The medical emergency submodule automatically obtains the coordinates of the nearest AED device through UWB positioning technology and directly connects to the hospital's emergency department, with a positioning error ≤ 0.5 meters. The specific steps are as follows: 1) When abnormal behaviors such as falls are detected, a health data request is triggered, and the passenger's identity identifier is quickly read through RFID / NFC tags; 2) Blockchain nodes verify identity and permissions, based on the Hyperledger-Fabric channel isolation mechanism, and smart contracts are automatically approved: rescue personnel need to pass two-factor authentication; 3) Decrypt and push health information. Data transmission uses TLS1.3 + national standard SM4 encryption, with a latency of ≤500ms; 4) After the emergency response is completed, the event record is written to the blockchain, which includes timestamp, location coordinates, data summary, and response result.
[0023] Preferably, the fire-fighting submodule includes a lidar fire source location unit, which can generate three-dimensional fire data and link with the fire pipeline control module. The fire location accuracy reaches ±15°. After a fire occurs, it automatically shuts off the power in the fire area and generates three-dimensional fire location data.
[0024] Preferably, it also includes a special passenger service module, which contains a biometric comparison unit that matches the identity information of special passengers through facial recognition.
[0025] Example 1 Taking a passenger fainting in a waiting hall as an example, high-definition positioning cameras distributed throughout the station hall continuously capture images of the waiting hall. An AI video analysis module identifies the fainting passenger and determines their posture by detecting key points on the body, such as the head, shoulders, elbows, and knees. When an abnormal posture, such as falling or lying flat, is detected and there is no significant movement for a period of time, an alert is triggered. Visual algorithms achieve seamless cross-area tracking and three-dimensional coordinate positioning of the target, outputting the type of abnormal behavior and its coordinates. Then, a health data request is triggered, quickly reading the passenger's identification identifier via RFID tags and confirming the passenger's identity through a biometric comparison unit. Next, the coordinates of the nearest AED device are automatically obtained through UWB positioning technology and directly connected to the hospital emergency department to contact rescue personnel. Based on the Hyperledger-Fabric channel isolation mechanism, the smart contract automatically approves the application and performs two-factor authentication on the rescue personnel by recognizing their facial features and identification badge information. Subsequently, the passenger information is decrypted to understand the passenger's past chronic disease history, and the data is encrypted using TLS1.3 + national cryptographic SM4 and pushed to the rescue personnel. The rescue personnel can then take emergency measures for the passenger as soon as possible based on the passenger's past medical history. Finally, after the emergency response is completed, the time of the incident, the coordinates of the incident location, the passenger data summary, and the results of the incident response are written into the blockchain record.
[0026] For customers' past medical history information, AES-256 encryption technology is used to encrypt and store personnel information, and a 24-hour automatic deletion mechanism based on timestamps is set up to prevent passenger information from being leaked.
[0027] Example 2 Taking the occlusion of a moving passenger target as an example, high-definition positioning cameras distributed in the station hall generate monitoring information and monitor the passenger's movement path. When the video surveillance detects that the target is occluded, the spatiotemporal context model is built in the LSTM network to predict the target's trajectory. The UWB positioning module continuously provides the coordinates of the surrounding environment anchor points. The Kalman filter algorithm is used to fuse the visual prediction results with the real-time UWB data to generate a compensated continuous coordinate stream. During the monitoring process, if an emergency occurs, such as a passenger fainting or other abnormal behavior, the steps in Example 1 can be repeated to carry out emergency rescue for the passenger. This end-to-end approach from abnormal behavior identification to emergency response reduces delays and enables rapid rescue operations.
[0028] Example 3 Taking a low-light environment at night as an example, high-definition positioning cameras distributed in the station hall generate monitoring information. Due to the low light intensity in the area, the multi-camera collaborative positioning module includes an infrared thermal imaging unit. In low-light environments, the target is continuously tracked through thermal radiation characteristics, with a thermal imaging resolution of 384×288 pixels. Furthermore, a multispectral fusion algorithm can be added, referencing Russian infrared thermal imaging technology, combining data from multiple sensors such as visible light, infrared, and lidar to improve the robustness of target recognition, enhance the clarity of the monitoring video in low-light environments at night, and improve the recognition effect.
[0029] Example 4 Taking a platform fire as an example, high-definition positioning cameras distributed in the station hall generate monitoring information. Through the lidar fire source positioning unit, three-dimensional fire data can be generated and linked with the fire pipeline control module. The UWB positioning module continuously provides the coordinates of anchor points in the surrounding environment. By using the Kalman filter algorithm to fuse visual prediction results with UWB real-time data, the fire positioning accuracy reaches ±15°. After the fire occurs, the power supply in the fire area is automatically turned off and three-dimensional fire positioning data is generated.
[0030] The above embodiments can be applied not only to emergency response management within stations, but also to other locations requiring coordinated management, such as schools, shopping malls, and subways. By proactively capturing abnormal passenger behavior, even if a passenger is unconscious and unable to call for help, the system can detect the abnormality immediately, display the specific location, and initiate rescue immediately, achieving "preemptive response." Furthermore, while issuing an early warning, the system will simultaneously push the passenger's medical history information, allowing rescue personnel to have a more comprehensive understanding of the situation and take targeted emergency measures, avoiding the impact of missing information on treatment effectiveness. Compared to traditional indiscriminate rescue, this approach is more precise and efficient.
[0031] All standard parts used in this invention can be purchased from the market, and irregular parts can be customized according to the description and drawings. The specific connection methods of each part adopt conventional methods such as bolts, rivets, and welding that are mature in the prior art. The machinery, parts and equipment adopt conventional models in the prior art, and the circuit connection adopts conventional connection methods in the prior art, which will not be described in detail here.
[0032] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.
Claims
1. A smart station emergency joint defense system, characterized in that, include: The multi-camera collaborative positioning module consists of distributed high-definition positioning cameras. It uses visual algorithms to achieve seamless cross-regional tracking and three-dimensional coordinate positioning of targets, with a positioning accuracy error of ≤0.3 meters. The AI video analysis module uses deep learning algorithms to identify abnormal behaviors such as falls, gatherings, and leaving items behind by people in the station in real time, and outputs the type of abnormal behavior and the coordinates of its occurrence. The emergency response module includes a digital emergency plan library, which automatically triggers associated fire alarms, power control, and evacuation guidance linkage operations when responding to abnormal events. The data security module uses AES-256 encryption technology to encrypt and store personnel information, and sets up a 24-hour automatic deletion mechanism based on timestamps.
2. The intelligent station emergency joint defense system according to claim 1, characterized in that: The multi-camera collaborative positioning module includes an infrared thermal imaging unit, which continuously tracks targets through thermal radiation characteristics in low-light environments. It can also add a multispectral fusion algorithm to combine data from multiple sensors such as visible light, infrared, and lidar to improve the robustness of target recognition.
3. The intelligent station emergency joint defense system according to claim 1, characterized in that: The multi-camera collaborative positioning module also includes a trajectory prediction unit, which predicts potentially dangerous behaviors by analyzing personnel movement paths.
4. The intelligent station emergency joint defense system according to claim 1, characterized in that: The AI video analysis module integrates an occlusion compensation algorithm, which continuously tracks the target when it is occluded through spatiotemporal context association. The specific steps are as follows: 1) Video surveillance detects that the target is occluded, and predicts the target's trajectory through spatiotemporal context modeling; 2) The UWB positioning module continuously provides the coordinates of anchor points in the surrounding environment, and uses the Kalman filter algorithm to fuse visual prediction results with real-time UWB data; 3) Generate a compensated continuous coordinate flow.
5. The intelligent station emergency joint defense system according to claim 1, characterized in that: The emergency response module includes a medical emergency submodule. This submodule automatically obtains the coordinates of the nearest AED device using UWB positioning technology and directly connects to the hospital's emergency department. The specific steps are as follows: 1) When abnormal behaviors such as falls are detected, a health data request is triggered, and the passenger's identity identifier is quickly read through RFID / NFC tags; 2) Blockchain nodes verify identity and permissions, and smart contracts automatically approve: Rescue personnel must pass two-factor authentication; 3) Decrypt and push health information; data transmission uses TLS 1.3 + SM4 encryption (Chinese national standard). 4) After the emergency response is completed, the event record is written to the blockchain, which includes timestamp, location coordinates, data summary, and response result.
6. The intelligent station emergency joint defense system according to claim 1, characterized in that: The emergency response module includes a medical first aid submodule and a fire handling submodule. The fire handling submodule includes a lidar fire source location unit, which can generate three-dimensional fire data and link with the fire pipeline control module. After a fire occurs, it automatically shuts off the power in the fire area and generates three-dimensional fire location data.
7. The intelligent station emergency joint defense system according to claim 1, characterized in that: It also includes a special passenger service module, which contains a biometric comparison unit that matches the identity information of special passengers through facial recognition.